Turtles make sense after all

Turtles may be weird, but according to new research, they’re not that weird. Their funny arrangement of shell and shoulder is just the same old land-dweller vertebrate stuff — with a little fold.

At first a turtle embryo grows much like a chicken or mouse. But then the developing body wall makes a critical fold, and the usual body plan starts to become an unusual turtle, Hiroshi Nagashima of Kobe University and his colleagues report in the July 10 Science.

Nothing else has a body plan like a turtle. Its ribs don’t grow inside its chest as a cage but instead fuse in the developing skin layer on its back to create one bony armored covering.

“It is not just that turtles ‘grew a shell,’” says paleontologist Ben Kear of La Trobe University in Melbourne, Australia. In the evolution of that shell, bones and muscles had to shift around relative to other reptiles, birds and mammals, and turtle shoulders ended up inside the rib cage. “In essence this means that the turtle skeleton is inside-out,” he says.

Odd as they are, turtles clearly belong to the lineage of amniotes, which includes mammals, birds and reptiles. Turtles, which are at least 200 million years old, “have survived all kinds of stuff — we’re talking extinction of the dinosaurs and myriad climate changes,” Kear says. Yet there’s scant fossil evidence of turtles in the making to explain how their forms arose as they split off from birds and crocodiles.

Knowing how a basic amniote embryo ends up developing into something so radically different could shed light on turtle history, says paleontologist Michael Lee of the South Australian Museum in Adelaide. “Some intermediate stages in this process might resemble real intermediate — fossil — stages in evolution,” he says.

To sort out how turtles develop, Nagashima and his colleagues worked with eggs of Chinese soft-shelled turtles (Pelodiscus sinensis) bought from a farm. The researchers used tissue-specific stains as well as substances that detect activity of particular genes to figure out which bits of the tiny embryos were on their way to becoming the bones and muscles of the adult. At each stage in development, the researchers compared their embryos with developing chickens and mice at comparable stages.

Any features shared by all three embryos probably came from distant common ancestors of all amniotes, including people, the researchers note.

In turtles, chickens and mice, the earliest stages of development looked much the same, the researchers reported. Then the turtle embryos veered off on their own path. The developing muscle tissue that would lie along adult ribs in a standard amniote began to fold underneath itself in the turtle. This tissue tucked inward, bending up to lie below the developing ribs. On this kinked-under section, the shoulder blades, or scapulas, formed.

If this fold could be straightened out, the scapulas would lie outside the rib cage, as they do in chickens, mice and people. For turtles then, “the position of the scapula is not a novelty,” Nagashima says. Essentially, “turtles have the same body plan as other amniotes.”

That critical fold in the tissue maps out the line that becomes an important embryological feature of turtle embryos called the carapacial ridge. Earlier research has shown that this ridge drives the development of the bony back of the animal. The fold also allows developing muscles to form connections in ways that they don’t in the mouse and the chicken.

The researchers also noted that the turtle ribs stop short in comparison with mice and chickens. Turtle ribs grow out only along the sides of what will become the backbone instead of curving into the body wall to form the whole rib cage. Those short turtle ribs mingle with the skin tissue creating the fused bony shell on the turtle’s back.

“Very, very sophisticated work,” says reptile paleontologist Olivier Rieppel of the Field Museum in Chicago in describing the extensive detective work required to trace all the tissues and muscles.

He has studied the oldest known fossil of an ancestral turtle, and he says the new interpretation of turtle embryology may fit well with the fossil record. Last year he and colleagues described Odontochelys semitestacea from a fossil collected in 220-million-year-old marine sediments in southwestern China. The turtle had a standard armored underside but not a full shell on its back. Its ribs  widened, but its shoulder blades still lay forward of instead of inside the ribs.

Nagashima speculates that the embryonic fold was evolving a bit at a time and maybe hadn’t reached as far around the body in this ancestral turtle as it does today. Clever suggestion, Rieppel says.

To understand turtle history, paleontologists really need more fossils, says Robert Reisz of the University of Toronto’s Mississauga campus in Canada. In the meantime, the new Japanese paper “clarifies a unique evolutionary event, one that gave rise to a really neat group of animals, our beloved turtles.”